Can the MMR or other vac ever cause autism?

[StarlightMcKenzie]

www.ncbi.nlm.nih.gov/m/pubmed/25114790/

Beachcomber - do you accept Hooker's claim about the data then, which essentially says that autism is not associated with MMR in any gender, any colour, any age, except for little black boys who were vaccinated between the ages of 24 and 36 months? Do you think that is the truth now?

No! Autism is not linked with the mmr .

I dont beleive it at all. Even if i did beleive it a little i would still vaccinate my babies. This may sound naive but measels mumps and rubella can be deadly, autism is not.

Catherina, why are you saying "vaccinated between the ages of 24 and 36 months" Didn't Hooker state in his results that there was a significant increase in autism cases specifically among African American males who received the first MMR prior to 36 months of age.

I don't see where he has ruled out MMR being associated with gender/age/race either. Can you point that out please? Also, can you comment on the figure you quoted earlier from LBRB as an estimate for how many black male children were included in the study. Are you happy with how Matt Carey came up with that figure?

Even without looking very hard I can see two major flaws with this paper.

1. Sample size - as others have said this is not given but given the data is split by sex, ethnicity AND age of vaccination none of the groups are going to be large and I would expect some to be very small indeed. As others have said it is highly suspicious that numbers in each group are not given as one reason would be to mask the small group size. The smaller the group size the more likely you are to get errors.

1. The other problem is that multiple tests have been carried out with no apparent statistical adjustment to take this into account. Put it like this if you do twenty tests on a set of data even if you take P

What Hooker has done is re-run CDC figures according to the original study's original untampered with protocol.

No he didn't. He did a cohort study, which is not the same as a case control study. You also saw from the stats link I gave you that a cohort study requires larger numbers than a case control study so using data that was collected for a case control study for a cohort study is already problematic.

The issue is not are Hooker's numbers big enough

Well it is an issue if you want to take his findings seriously isn't it. Except we don't know what his numbers were because he left them out of his tables, despite it being standard to include them. It smacks of intellectual dishonesty, if he is actually hiding small sample sizes by omission.

And surely if you are going to use your study as a launching board for an attack on the CDC and their intellectual dishonesty, you'd want your study to be open and transparent....if you had nothing to hide?

I was under the impression that the reason Hooker looked at that particular subgroup was because Thompson had mentioned that it had come up as significant in the original paper and that had 'hidden' it. I don't think he just went digging and playing around with the data for hell of it.

The original data set was also split by sex, ethnicity and age of vaccination. Do you think their groups were also too small to record anything significant?

I'm not sure OP but I think if you look hard enough you can find a link between any two things.

I was worried about my dcs having the mmr but then thought to myself that I was worrying about a risk that had not been proven in any way whereas the link between measles and terrible complications was proven so I made the best choice I could. Dcs were fine but I did worry initially I won't lie after a few weeks I was very happy that they were ok.

What size sample do you think you need for a cohort study noble?

Do you think their groups were also too small to record anything significant?

I don't know, I've not read the paper. It could be just as dodgy as this one, but at least I'd know, because the authors appear to have actually included their sample sizes.

Why didn't Hooker?

I have heard it said that medicine is not an exact science. Nobody can say MMR does not ever cause autism or contribue to the development of autism in a child. How can they. They can't.

www.tylervigen.com/

Correlation does not mean causation. Look here for evidence.

I don't know noble but I don't automatically assume the worst ie. that he was trying to hide small sample sizes because I don't want to agree with the findings. As I mentioned before, the majority of the birth cert sample were black(table 5) and the majority were male(table 3). The vast majority of the total sample were vaccinated before 36 months as well (table 1 - 93%) but I'm not sure what the breakdown was for age of vaccination in the birth cert sample. I don't yet know if that was given in the original paper. However, based on that information I would't automatically assume that the samples were tiny.

noble, you've mentioned a couple of times that you are concerned about the use of data for a case control study being used in a cohort study due to sample size. What size sample do you think you need for a cohort study?

Orange, yes, correlation does not automatically mean causation but it may suggest causation and may warrant further investigation.

Well leaving out the sample sizes means either something to hide or incompetence, neither of which should inspire confidence.

The question in the OP was can the MMR ever cause autism. From the paper provided which suggests yes, there isn't actually enough information given to decide whether it is good evidence or not. So that's a non-starter.

mrs, I don't know, but I did post a link to a stats website upthread that says a case-control study can run on smaller numbers and a cohort study requires larger numbers. So even if the subgroups in the case control study were small, that wouldn't be as much of an issue a in a cohort study and saying that any criticism levelled at the cohort study would automatically have to apply to the case-control study isn't necessarily true.

That said, I'm not a statistician, just going from what it said on the stats website.

In this case the cohort size was 624 cases and 1824 controls.

Yes, but the interesting finding was from an analysis run on a subgroup of the population! The subgroup was African American boys (without low birthweight). The size of that subgroup is the important thing.

You can't say 'we looked at over two thousand kids and we found out something really interesting about black kids' if there's only one black kid in the group. That is surely obvious.

im glad i found this thread and this discussion. i read about this paper recently and its had me wanting to learn more about this subject.

someone upthread mentioned lawsuits in the USA and i wanted to mention the case of Hannah Poling and her parents. they won a lawsuit against the US federal authorities for the vaccine programme making their daughter develop autism. specifically that the large series of vaccines they gave her daughter in one sitting, mmr one of them, led to her autism. i havent been able to find out the sum of the payout but it was awarded by the vaccine court, the U.S. Court of Federal Claims, "which administers a no-fault system for litigating vaccine injury claims." they conceded the case to her family without proof of causation.

content.time.com/time/health/article/0,8599,1721109,00.html

what made me interested in this particular study about the relationship between autism and mmr in young african american boys in georgia is that only last year authorities in Minneapolis USA conceded that there were especially high rates of autism in somali children (another african group) in their state. the study also found higher rates in white children compared to the national average too - is it environmental? the study didnt give a reason just noted the pattern. but is highly specific to that somali community in that state, not seen in other somali communities around the USA. furthermore the high rates of autism are in children born in the usa compared to somali children born in somalia who emigrated with their parents to the usa.

along with the studies from sweden that also saw higher rates of autism in the somali community there has been theories put forward about lack of vitamin d and link to autism. but, there are large somali immigrant groups in the uk of whom many are born in the uk, and autism rates in that subgroup are not higher than the majority demographic group even with similar low levels of sun light and vitamin d from dark skin and covering up from the sun.

but then theres another study showing high rates of autism in ethiopian children born in israel compared to children born in ethiopia whose families emigrated (of whom NON have autism). www.vaccinationnews.org/20110121AutismVaccinationImmigrantsYazbakFE (thats the only place i could find the israel studies) high sunshine in israel, different levels of dress covering because the ethiopian israelis are jews not muslims like the somalis, but still, higher rates of autism compared to in ethiopia where autism is seen as more prevalent in children of people working/travelling through the west . www.bbc.co.uk/news/world-africa-14991267

Actually, if you look at Table 1 you can see that there was a statistically significant effect for the entire cohort at 36 months.

Table 2 shows the effect on African American boys in the entire cohort. There were 500 boys in the entire cohort. In the birth-cert cohort, over 60% were black children. Even if you want to be more cautious with that figure when you look at the entire cohort, you are still talking about a couple of hundred at least in the black, male group in the entire cohort.

In the birth cert group there are 243 boys. Based on the 60% from above that's 145 black children. Only 49 children in the birth cert group fell into the low birth weight category.Even if you assume that they are all black boys (which is incredibly unlikely) there are still 96 children in that group. I'm just not seeing evidence of 'tiny samples'. Why choose to jump to that conclusion?

Obviously the above are all estimates but even if you are conservative(eg. all low birth weight children were black males), it doesn't imply small samples.

So you're taking the line that Hooker was incompetent then?

How does that suggest he is incompetent? I'm pointing out that it is unlikely, based on the data that was used, that the sample sizes are going to be tiny. Do you agree?

We are having this discussion about sample sizes because we don't know the sample sizes. We don't know the sample sizes because Hooker neglected to include them in his report, despite them being vital to assessing the importance of his results.

If he was competent and honest, he would have known they were vital and included them. They aren't included, so either his competency or his honesty are suspect.